Each year, the software group at IBM puts on an analyst event, this year in Stamford, Connecticut. All 4 of the RedMonks are here as its a good chance to get an update on what IBM’s done over the past year and, better, start to figure out what they’ll be doing. Thus far I’ve attended two sessions:
Advanced Analytics with Brenda Dietrich
Joined with the hype over Big Data (the potential to do more number crunching and analytics due to advances in technologies that make it cheaper and faster) I’m always curious what companies are actually using it for beyond the Web 2.0 examples. Starting with a basic overview of what “analytics” actually means in a business context (create a model, test a hypothesis that helps you decide something to change, feed the results back in the chain, repeat) Dietrich then went into some customer cases of Big Data analysis that were actually pretty grounded:
- The US Social Security Administration speeding up part of their paper-work process.
- New York State figuring out the best ways, per type of person, to collect delinquent taxes
- A South African cellphone company using analytics to study clusters of people (as determined by their call logs) and figure out why they cancel their subscriptions.
- A power company figuring optimizing the way it runs its grid based on analytics (I admit, I didn’t catch it fully).
- IBM sales running analytics over sales and accounts to make sure they were maxing out their customer budgets – collecting the maximum amount of cash they have for IT.
(On the topic of base lining what all this analytics hoopla is about, I enjoyed Milton Michael’s Data Analysis book – I’m one of those guys who took math for poets in college, so I need anything beyond addition spelled out real easy.)
Image management with David Linquist and Pratik Gupta
I was only in this session for the first half (due to a nice ColdFusion cloud update), but they jumped into the demo quickly and went over all the different storage back-ends for images. Of course, in the back of my head the whole time was the Big Foil Ball polemic. As Pratik Gupta said, managing operating systems is pretty easy, it’s those pesky applications and all the configuration that comes with them that cause the problems.
The cloud angle – the full session title was “The Next Wave of Image Management with Virtualizaton and Cloud Computing” – was using S3 (and, presumably, other cloud storage) to host the image library and also deploy images to behind-the-firewall or beyond. There was even some CMDB talk with one analyst asking how all this was accounted for CI-wise.
Sitting in the session, it’s clear that the Puppet, Chef, and other next generation automation/image management people still have their work cut out for them when it comes to traditional enterprises: there’s no end to the need for “the old way of doing it didn’t work, the new way does” case studies from that crowd.
The other open question with IBM is always around what their virtualization alliances are and will be. Here, image management worked with VMWare and KVM, and there was much hope and use around OVF, either using it straight up or extending it.
I hear there was some exciting Hadoop stuff from the IBM labs in another session. James was certainly excited about it. The rest of the event is a mix of general session – from Steve Mills, head of IBM Software, and others – and more break-out sessions.
What I’m looking for here is how IBM is going to motivate the world’s enterprises to spend money on all these new geegaws and technologies. We seem to be at another point in IT innovation where the features and functionality available are far ahead of what companies are asking for (or know to ask for). The stuff of cloud-nut daily frothing like advanced development and automation, for example, are too uncontrolled, unknown, and new at this point for companies to quantify the risks and benefits of use. Companies like IBM (and all its peers) along with the associated communities need to help IT (and their companies) transition to using these new technologies without slicing off too many toes in the process.
IBM’s done a good job over the past year creating the start for desire in companies – think about all those blue-barred ads and industry frameworks like SAFE (video above) – but there’s been an open question of how IBM can insert itself into scenarios like FedEx Critical’s truck tracking, an enterprise-class application that didn’t seem to involve IBM in a big way. That’s why examples from the analytics session above are interesting: they’re examples of IBM getting paid to help customers take advantage of new technologies instead of just managing older ones.